Datasets:
Tasks:
Feature Extraction
Modalities:
Image
Formats:
imagefolder
Languages:
English
Size:
< 1K
Tags:
medical-imaging
computational-pathology
digital-pathology
immunohistochemistry
foundation-model
pan-cancer
License:
| license: apache-2.0 | |
| task_categories: | |
| - feature-extraction | |
| language: | |
| - en | |
| tags: | |
| - medical-imaging | |
| - computational-pathology | |
| - digital-pathology | |
| - immunohistochemistry | |
| - foundation-model | |
| - pan-cancer | |
| - feature-representation | |
| - pytorch | |
| viewer: false | |
| # ImmunoBench Feature Representations | |
| ## π Overview | |
| <p align="left"> | |
| <b> | |
| π€ <a href="https://huggingface.co/datasets/AI4Pathology/ImmunoBench-image-features">Dataset</a> | | |
| π <a href="https://yanfang-research.github.io/ImmunoBench-Web/">Homepage</a> | | |
| π <i>Paper (coming soon)</i> | | |
| π» <a href="https://github.com/yanfang-research/ImmunoBench">Code</a> | |
| </b> | |
| </p> | |
| This repository provides **pre-extracted feature representations** from whole slide images (WSIs) from the **ImmunoBench** project, a large-scale benchmark designed for evaluating pathology foundation models in IHC-centered pathology workflows. ImmunoBench integrates data from 13,657 patients, 26,058 WSIs, 9,682,431 TMA-derived IHC images and 464,294 annotated patches, collected across 13 institutions, covering 27 organs, 43 clinical stains and 14,708 gene-encoded proteins. | |
| <table> | |
| <tr> | |
| <td align="center" width="50%"> | |
| <img src="assets/dataset_circos.png" width="95%"> | |
| </td> | |
| <td align="center" width="50%"> | |
| <img src="assets/task_sunburst.png" width="95%"> | |
| </td> | |
| </tr> | |
| <tr> | |
| <td align="center"> | |
| <b>Figure 1.</b> ImmunoBench dataset overview | |
| </td> | |
| <td align="center"> | |
| <b>Figure 2.</b> ImmunoBench task taxonomy | |
| </td> | |
| </tr> | |
| </table> | |
| --- | |
| ## π Access Policy | |
| Access to this dataset is **restricted** and subject to approval. | |
| To request access: | |
| - Please contact the dataset maintainers using your **institutional email address** | |
| - Clearly state your **affiliation, research purpose, and intended use** | |
| > β οΈ This dataset is provided **for academic research only**. | |
| > Commercial use is **strictly prohibited** without prior written permission. | |
| --- | |
| ## π Dataset Versioning & Updates | |
| This dataset is part of an actively maintained benchmark and will be **continuously updated**. | |
| - Current release: `HANCOCK` subset | |
| - Future releases may include: | |
| - Additional cohorts | |
| - More foundation models | |
| - Extended feature modalities | |
| All updates will be documented in this repository and the project homepage: | |
| π [Project Homepage](https://yanfang-research.github.io/ImmunoBench-Web/) | |
| --- | |
| ## β οΈ Usage Notes | |
| - These features are **pre-extracted representations**, not raw images | |
| - Performance may vary depending on downstream task design | |
| - Users are responsible for ensuring compliance with local data regulations | |
| --- | |
| ## π€ Contributions | |
| We welcome community contributions and collaborations. | |
| If you would like to propose additional datasets, tasks, or benchmark extensions for inclusion in **ImmunoBench**, please contact us or open an issue in our GitHub repository. | |
| π» [GitHub Issues](https://github.com/yanfang-research/ImmunoBench/issues) | |
| --- | |
| ## π« Contact | |
| For access requests, collaboration inquiries, or technical issues: | |
| π§ **test@pjlab.org.cn** | |
| --- | |
| ## π Citation | |
| If you use this dataset in your research, please cite: | |
| [*Citation information coming soon.*] | |
| --- |